Integrating AI and Biometric Information for Personalized Alcohol Intervention: The Step Away Alcohol Manager (SAAM) - PROJECT SUMMARY / ABSTRACT Hazardous alcohol consumption remains a significant public health issue in the U.S., responsible for over 140,000 deaths annually. Despite its profound impact, only about 10% of individuals diagnosed with alcohol use disorder (AUD) receive any form of treatment, underscoring a substantial treatment gap. This SBIR Phase I proposal seeks to address this gap by enhancing the Step Away Alcohol Manager (SAAM), a mobile health (mHealth) intervention developed by Here and Now Systems, LLC (HNS). SAAM builds on the success of the Step Away app, which has demonstrated significant reductions in alcohol use —50-60% reductions in four pilot trials. However, sustaining long-term user engagement remains a challenge, with only 40% of users continuing to engage with the app after six months and some modules being used infrequently. This project aims to integrate cutting-edge artificial intelligence (AI) and biometric technologies to enhance personalization, improve engagement, and increase the app’s overall effectiveness in promoting long-term behavior change. Specifically, we will incorporate a multi-agent generative AI (GenAI) and large language model (LLM) to provide users with empathic, real-time conversations and strategies to manage drinking-related challenges such as difficulty sleeping, stress, and cravings. The AI will initiate interactions during users' high-risk times for drinking and during daily check-ins, providing proactive and responsive support. Additionally, we will integrate facial scan technology which will monitor users’ stress levels, sleep quality, and overall health in real time. This biometric data will be used to tailor feedback and enhance the relevance of the AI-driven interactions, further personalizing the user experience. The Phase I project has three specific aims: (1) Develop SAAM with advanced AI-driven conversational capabilities and real-time biometric feedback. The chatbot will deliver evidence-based responses based on trusted sources, ensuring it adheres to scientifically sup ported interventions. (2) Conduct end-user testing with 20 participants who meet AUDIT criteria for hazardous drinking to evaluate the app’s content, usability, and the quality of the personalized messages generated by the A I. We will use validated tools such as the Client Satisfaction Questionnaire (CSQ) and System Usability Scale (SUS), with benchmarks of 24+ on the CSQ and 70+ on the SUS. We will also utilize qualitative research methodology to determine themes related to SAAM’s strengths and limitations, which will be used to inform needed modifications. (3) Refine SAAM based on feedback from these reviews, addressing any usability issues, content gaps, and engagement barriers to create a polished, user-friendly app ready for Phase II evaluation. Through this Phase I project, SAAM has the potential to significantly expand access to effective, evidence-based interventions for alcohol use disorder by leveraging advanced AI and biometric technologies to provide real-time, personalized support. This innovative approach will address the critical public health need for more accessible and engaging alcohol treatment solutions, offering a novel, scalable intervention with the potential for widespread adoption.
